Reboot of MOS 8 MU-3 controller node returns it to bootstrap state

[Rudy McComb]

The controller nodes in my environment were rebooted during a failover test and when they returned online they were set back to the bootstrap mode. The solution for this problem is to set the "netboot-enabled" flag to "false" inside the cobbler system settings for the cobbler container. This can be accomplished by running the following command for each affected node that experienced this unexpected behavior.

1. Disable PXE boot behavior of the system in Cobbler:

# dockerctl shell cobbler cobbler system edit --name=<insert_name_of_your_node> --netboot-enabled=false

2. Confirm that the changes have taken place correctly:

#dockerctl shell cobbler cobbler system report --name=<insert_name_of_your_node> | grep -i "Netboot enabled"

Has anyone else experienced this in an MOS 8.0 MU3 patched environment?

[Vladimir Kozhukalov (kozhukalov)]

Could you please provide us a diagnostic snapshot from the Fuel master node? And please also describe the exact sequence of steps that could be used to reproduce the issue.

[Rudy McComb (rmccomb)]

Here is the snapshot from the Fuel Master.
All it took was a reboot of any node and it would end up in bootstrap mode instead of the role it was deployed in.